Eye on AI

→ WHAT IT COVERS Alexander Whedon, CTO of SubQuadratic, explains how his company's sparse attention mechanism (SSA) reduces transformer compute costs by 40x at 1 million tokens, enabling multimillion-token context windows that reshape enterprise RAG architectures, coding agents, and long-horizon AI workflows without quadratic scaling penalties. → KEY INSIGHTS - **Coding agent efficiency:** Benchmarking frontier models on SWEBench Pro revealed that 86% of agent steps are read operations —...

→ WHAT IT COVERS Marko Kushnir, communications director at General Cherry, Ukraine's top-five drone manufacturer, details how the company scaled from 10 drones monthly in 2023 to nearly 100,000, covering AI targeting systems, fiber optic drones up to 40km range, interceptor technology destroying Russian Shahed drones at a 70-80% success rate, and post-war commercial expansion plans.

→ WHAT IT COVERS Yaroslav Azhnyuk, co-founder of The Fourth Law, details how Ukraine's active warzone has become the world's primary autonomous drone development environment, outlining a five-level autonomy framework for weapons systems and arguing that AI-guided weapons will become an ethical necessity within a decade. → KEY INSIGHTS - **Five-Level Drone Autonomy Framework:** Azhnyuk's classification runs from Level 1 (autonomous terminal guidance after human targeting) through Level 5 (fully...

→ WHAT IT COVERS Alexander Palamarchuk, a drone R&D specialist with Ukraine's Azov Corps operating 18 kilometers from the front line, details how both sides use AI for target recognition, autonomous navigation systems, jamming countermeasures, and the trajectory toward fully autonomous lethal drone swarms within two to three years. → KEY INSIGHTS - **Jamming countermeasures:** Ukrainian drone units have expanded operational radio frequencies from standard 2.

→ WHAT IT COVERS Manoj Saxena, CEO of Trustwise, explains why 95% of AI agent projects fail to reach production and presents the case for a dedicated runtime control layer — called an AI control tower — that governs multi-agent systems across safety, compliance, and cost dimensions simultaneously, drawing on deployments at banks, insurers, and global brands. → KEY INSIGHTS - **The 95% Production Gap:** MIT research confirms that 95% of AI agent projects stall between pilot and production.

→ WHAT IT COVERS Foundation Future Industries cofounder Mike LeBlanc describes building 150 humanoid robots in 20 months, deploying them in an automotive factory in Georgia on 24-hour shifts, pursuing Department of Defense contracts across Army, Navy, Air Force, and Marine Corps, and planning battlefield testing in Ukraine — while host Craig Cannon maintains skepticism about production claims and technical readiness.

→ WHAT IT COVERS Sid Sheth, CEO of d-Matrix, explains how a premium token economy is emerging around low-latency AI inference, where users pay 10x more ($20 vs $2 per million tokens) for real-time interactivity. He covers d-Matrix's chiplet architecture, the Giga IO acquisition, agentic enterprise adoption, and sovereign AI infrastructure demand. → KEY INSIGHTS - **Premium Token Pricing:** Two distinct inference tiers now exist in the market.

→ WHAT IT COVERS Drew Cukor, former Marine colonel and JPMorgan AI transformation lead, now at TWG AI, argues that legacy American enterprises have 36 months to transition from Microsoft Office-dependent workflows to AI-native platforms using structured data layers and Palantir Foundry, or face displacement by AI-native competitors and Chinese firms.

→ WHAT IT COVERS Paolo Ardoino, CEO of Tether, argues that within five years, 90% of everyday AI use cases will run locally on smartphones, making massive centralized data centers largely redundant for consumer applications, while explaining Tether's QVAC platform and its open-source approach to edge AI deployment. → KEY INSIGHTS - **Edge AI Timeline:** By 2029, smartphones will handle 70% of common AI tasks locally; by 2031, that figure reaches 90%.

→ WHAT IT COVERS Sid Sheth, CEO of d-Matrix, explains how a premium token economy is emerging around low-latency AI inference, where users pay 10x more ($20 vs $2 per million tokens) for real-time interactivity. He covers d-Matrix's chiplet architecture, the Giga IO acquisition, agentic enterprise adoption, and sovereign AI infrastructure demand. → KEY INSIGHTS - **Premium Token Pricing:** Two distinct inference tiers now exist in the market.

→ WHAT IT COVERS Erhan Giral from BMC Helix explains how agentic AI is transforming IT service management by autonomously detecting anomalies, performing root cause analysis, and generating remediation plans for Fortune 2000 enterprises — shifting IT operations from reactive firefighting to automated, self-improving systems that learn from every resolved incident.

→ WHAT IT COVERS Chris Blackburn, founder and CEO of Liatrio, explains why genuine AI transformation costs the equivalent of every employee's salary spread across two years, why most enterprises claiming AI adoption are nowhere near transformed, and how the "strangler fig" organizational pattern offers a practical path forward for large, complex organizations.

→ WHAT IT COVERS Dr. Kate Adamala, synthetic cell engineer at the University of Minnesota, explains how her lab created "spud cells" — synthetic lipid vesicles containing purified biological molecules capable of genetically encoded feeding and division, representing a milestone toward engineering biology as a general-purpose manufacturing technology for medicine and climate solutions.

→ WHAT IT COVERS Victor Riparbelli, CEO of Synthesia, explains how the company pivoted from Hollywood ambitions to enterprise training software, and now launches a two-way interactive avatar product targeting Fortune 100 companies, with live AI agents replacing one-way corporate video communication across sales training, HR, and education. → KEY INSIGHTS - **Enterprise market fit over Hollywood:** Synthesia found its first viable market not with film studios but with corporate training teams...

→ WHAT IT COVERS Dr. Kate Adamala, synthetic cell engineer at the University of Minnesota, explains how her lab built "spud cells" — synthetic lipid vesicles containing purified biological molecules capable of genetically encoded feeding and division, representing a platform to make biology a general-purpose manufacturing technology for medicine and climate solutions.

→ WHAT IT COVERS Virtana CEO Paul Appleby presents findings from the AI Factory Reality Check study, revealing that 6 in 10 enterprises cannot automatically identify root causes when AI workloads fail, while governance investment lags behind infrastructure spending across banking, healthcare, airlines, and retail sectors deploying on-premises AI data centers.

→ WHAT IT COVERS Bradon Rogers, Chief Customer Officer at Island, explains how the company's enterprise browser addresses AI security risks by wrapping policy controls around consumer AI tools, agentic workflows, and MCP calls — enabling organizations to empower users with AI while maintaining data protection, regulatory compliance, and audit trails across all devices.

→ WHAT IT COVERS Kriti Sharma, CEO of IFS's Nexus Black unit, explains how her team deploys production-grade industrial AI across manufacturing, utilities, and aerospace by embedding engineers on-site at refineries, hangars, and factories to solve high-stakes operational problems within three weeks of engagement. → KEY INSIGHTS - **On-site deployment model:** Nexus Black achieves measurable client value within three weeks by physically embedding engineers at industrial sites — refineries,...

→ WHAT IT COVERS Devvret Rishi, CEO of PredaBase (acquired by Rubrik), explains why AI agents represent the next major enterprise security threat vector, how Rubrik's Agent Cloud platform governs agents across multi-vendor environments, and why organizations are stuck between blocking AI entirely or deploying it without adequate risk controls. → KEY INSIGHTS - **Agent Definition & Risk Scope:** Enterprises should treat any model with access to tools, APIs, or databases as an agent requiring...

→ WHAT IT COVERS Vin Singh, founder and CEO of BulFrog AI (Nasdaq: BFRG), explains how his three-platform AI system — BF Prep, BF Leap, and BF Arenas — targets the 50% Phase 3 clinical trial failure rate in pharma by improving data quality, causal gene discovery, and drug target selection. → KEY INSIGHTS - **Phase 3 Failure Economics:** Big pharma fails 50% of Phase 3 clinical trials despite each drug requiring 10–15 years and $1–2 billion in development.